pyVideoTrans

Appearance

Local Large Language Models and Usage of Domestic AI Compatible with OpenAI ChatGPT Interface ​

In video translation and dubbing software, large language models can serve as efficient translation channels, significantly improving translation quality by leveraging context.

Currently, most domestic AI interfaces are compatible with OpenAI's technology, so users can operate directly in OpenAI ChatGPT or local large language models. Local deployment using ollama is also possible.

Moonshot AI Usage ​

  1. Fill in the API address with https://api.moonshot.cn/v1
  2. Fill in the SK with the API Key obtained from the Moonshot open platform, which can be obtained from this URL: https://platform.moonshot.cn/console/api-keys
  3. Fill in the model text box with moonshot-v1-8k,moonshot-v1-32k,moonshot-v1-128k
  4. Then select the desired model in the model selection area. Keep it after testing without problems.

DeepSeek AI Usage ​

  1. Fill in the API address with https://api.deepseek.com/v1
  2. Fill in the SK with the API Key obtained from the Moonshot open platform, which can be obtained from this URL: https://platform.deepseek.com/api_keys
  3. Fill in the model text box with deepseek-chat
  4. Then select deepseek-chat in the model selection area. Keep it after testing without problems.

Zhipu AI BigModel Usage ​

  1. Fill in the API address with https://open.bigmodel.cn/api/paas/v4/
  2. Fill in the SK with the API Key obtained from the Moonshot open platform, which can be obtained from this URL: https://www.bigmodel.cn/usercenter/apikeys
  3. Fill in the model text box with glm-4-plus,glm-4-0520,glm-4 ,glm-4-air,glm-4-airx,glm-4-long , glm-4-flashx ,glm-4-flash
  4. Then select the desired model in the model selection area. You can choose the free model glm-4-flash. Keep it after testing without problems.

Baichuan Intelligent AI Usage ​

  1. Fill in the API address with https://api.baichuan-ai.com/v1
  2. Fill in the SK with the API Key obtained from the Moonshot open platform, which can be obtained from this URL: https://platform.baichuan-ai.com/console/apikey
  3. Fill in the model text box with Baichuan4,Baichuan3-Turbo,Baichuan3-Turbo-128k,Baichuan2-Turbo
  4. Then select the desired model in the model selection area. Keep it after testing without problems.

Notes: ​

  1. Most AI translation channels may limit the number of requests per minute. If an error message indicates exceeding the request frequency, click "Translation Channel ↓" on the main software interface, and in the pop-up window, change the pause seconds to 10. This means waiting 10 seconds after each translation before initiating the next translation request, allowing a maximum of 6 times per minute to prevent exceeding the frequency limit.

  2. If the selected model is not intelligent enough, especially locally deployed models limited by hardware resources, which are usually smaller and unable to accurately return translations in the required format according to instructions, there may be too many blank lines in the translation results. In this case, try using a larger model or open Menu--Tools/Options--Advanced Options--Send complete subtitle content when using AI translation, and uncheck it.

Using ollama for Local Deployment of Tongyi Qianwen Large Language Model ​

If you have some hands-on ability, you can also use this model for translation after local deployment. This section uses Tongyi Qianwen as an example to introduce deployment and usage.

1. Download the exe and Run Successfully ​

Open the URL https://ollama.com/download

Click to download. After downloading, double-click to open the installation interface and click Install to complete.

A black or blue window will automatically pop up after completion. Enter three words ollama run qwen and press Enter. This will automatically download the Tongyi Qianwen model.

Wait for the model download to finish. No proxy is needed, and the speed is quite fast.

After the model is automatically downloaded, it will run directly. When the progress reaches 100% and displays the "Success" character, it means the model has run successfully. This also indicates that the Tongyi Qianwen large language model installation and deployment are complete, and you can start using it happily. Isn't it super simple?

The default API address is http://localhost:11434

If the window is closed, how do you reopen it? It's also simple. Open the computer's start menu, find "Command Prompt" or "Windows PowerShell" (or directly type Win key + q key and input cmd to search), click to open it, and enter ollama run qwen to complete.

2. Use Directly in the Console Command Window ​

As shown in the figure below, when this interface is displayed, you can actually start typing text in the window to use it.

3. Of Course, This Interface Might Not Be Very User-Friendly, So Let's Get a User-Friendly UI ​

Open the URL https://chatboxai.app/zh and click download

Double-click after downloading and wait for the interface window to open automatically.

Click "Start Setting". In the pop-up layer, click "Ollama" in the top model and AI model provider, fill in the API domain name address http://localhost:11434, and select Qwen:latest in the model drop-down menu, then save and you're done.

The usage interface is displayed after saving. Use your imagination and use it freely.

4. Fill in the API into the Video Translation and Dubbing Software ​

  1. Open Menu--Settings--Local Compatible OpenAI Large Model, add a model ,qwen in the middle text box, as shown below, and then select this model

  1. Fill in http://localhost:11434 in the API URL, and fill in SK arbitrarily, for example, 1234

  1. Test whether it is successful. Save it if successful, and then use it.

5. Other Usable Models ​

Besides Tongyi Qianwen, many other models can be used. The method is just as simple. You only need three words ollama run model name.

Open this address https://ollama.com/library to see all model names. If you want to use one, copy the name and then execute ollama run model name.

Remember how to open the command window? Click the start menu and find Command Prompt or Windows PowerShell.

For example, if I want to install the openchat model

Open Command Prompt, enter ollama run openchat, press Enter, and wait until "Success" is displayed.

Notes: ​

  1. Most AI translation channels may limit the number of requests per minute. If an error message indicates exceeding the request frequency, click "Translation Channel ↓" on the main software interface, and in the pop-up window, change the pause seconds to 10. This means waiting 10 seconds after each translation before initiating the next translation request, allowing a maximum of 6 times per minute to prevent exceeding the frequency limit.

  2. If the selected model is not intelligent enough, especially locally deployed models limited by hardware resources, which are usually smaller and unable to accurately return translations in the required format according to instructions, there may be too many blank lines in the translation results. In this case, try using a larger model or open Menu--Tools/Options--Advanced Options--Send complete subtitle content when using AI translation, and uncheck it.